It is most probable that both genetic and environmental factors contribute to colon cancer etiology. At present, we have the capabilities to look at the interaction between dietary intake and genes that have been linked to colon cancer. In this study, we will evaluate DNA obtained from tumor tissue from cases enrolled in a large population-based case-control study of colon cancer (CA61757; CA 48998); we propose to collect tumor blocks from a population-based case-control study of rectal cancer to further evaluate associations (continuation of CA48998). Tumor DNA will be analyzed to determine specific mutations and CIMP pathway phenotype. We will evaluate the associations between these genetic mutations and tumor stage at diagnosis and survival. These data will be linked to environmental data, that includes extensive information on dietary intake within the population, to determine the impact that diet has on causing these somatic mutations (location and type of mutation). Total calories, fat, protein, calcium, fiber, beta carotene, and folic acid will be assessed with these genetic mutations, as will meats (along with method of and degree of cooking), dairy products, legumes, soy products, and fruits and vegetables. Other factors such as physical activity and body size which are closely related to dietary intake will be assessed both for their associations with somatic mutations. It is hypothesized that dietary intake (as specified above), physical inactivity, and a larger body size will contribute to the CIMP phenotype. Using data from the original studies, we will use statistical methods to better define disease pathways. We will include previously collected information on p53, K-ras, and microsatellite instability in tumors as well as information on CIMP phenotype. Rectal tumors will be characterized in the same manner as colon cancer tumors were originally characterized (i.e. p53, K-ras, and microsatellite instability). Differences in colon and rectal tumors will be compared. Additionally data from rectal tumors will be combined with that from colon tumors to define disease pathways.